This week, Altair Engineering announced it has opened up its HyperWorks CAE platform to include third-party applications. The company is adding 15 new ISV applications on top of the existing 28 native HyperWorks titles. The new offerings are spread across six ISV partners, who have developed versions of their software that will run under the HyperWorks on-demand model.
In this case, the on-demand aspect refers to the licensing model rather than a hosted software environment. Altair has implemented a unique licensing scheme, wherein a range of applications can share a common pool of license tokens. When an application is fired up, it draws tokens out of the pool; when it’s done, the tokens are returned to the pool for other applications to use. By only charging for concurrent use, customers have more control over how they buy and use software. It’s especially useful for global enterprises that operate across multiple time zones or for businesses that have a lot of overlapping peaks and valleys in their software usage. Compared to more traditional licensing approaches based on the underlying hardware platform, Altair’s model is closer to the ideal of pay-per-use.
For current HyperWorks users that have not saturated their token usage, they will theoretically be able to run these new third-party applications at zero cost. In that scenario, taking advantage of the additional offerings is a no-brainer. Even if the tokens are maxed out a lot of the time, it still might make sense to access the other applications under HyperWorks if the software is to be used infrequently. Users would just need to purchase additional tokens and rebalance their workloads.
On the other hand, if a customer is using a third-party application a lot, it probably makes sense to deal directly with the vendor. Altair’s partner pricing structure is set up so that customers pay a premium for the “on-demand” access. (Altair sends monthly royalty checks to their partners based on a pretty simple formula that is derived from actual software usage.) But if the user knows that they will be running a given application the majority of the day, it’s probably better to go back to the original ISV and buy in bulk.
“It complicates the decision for the customer to the extent they have to understand their usage pattern,” admits Michael Humphrey, vice president of the Altair Partner Programs. But he says the company offers some tools to help them do that.
Humphrey believes that over the next year, Altair is probably going to be subsidizing the program to pay royalties to the partners. But over time, they expect that users will increase their overall software usage and saturate their token pool — so they’ll have to purchase more. But from the customer point of view, they are getting better value by using their tokens more completely. Their incremental cost is less because the overall utilization is better.
Altair is not doing this out of altruism. By offering third-party applications under its licensing umbrella, Altair is attempting to elevate HyperWorks from a being just a suite of CAE tools to a platform for a much wider market. HyperWorks itself is especially strong in pre-and post-processing, FEA, multibody dynamics, crash analysis, and static stress/strain applications. The six new partners bring in CFD, electromagnetics, business intelligence/analytics, and fatigue/durability codes. Here’s the breakdown:
ECS-Magna Powertrain: FEMFAT, a fatigue/durability suite.
EMSS: FEKO, an electromagnetics suite.
HiQube: a business intelligence & analytics suite.
Metacomp Technologies: MIME (Multipurpose Intelligent Meshing Environment); CFD++, a CFD suite; and CAA++, a computational aeroacoustics software suite.
nCode: DesignLife, a fatigue/durability analysis suite.
Software Cradle: SC/Tetra, another CFD suite.
According to Humphrey, they basically cherry-picked the software they felt best complemented their in-house HyperWorks titles. The immediate goal is to offer stiffer competition to Altair’s biggest competitors, MSC.Software and ANSYS. Both rivals offer a much wider range of titles than Altair, a lot of which were collected through acquisitions or OEM reseller arrangements.
ANSYS, in particular, has been especially busy buying up smaller ISVs to build up its stable of engineering codes. Over the last several years, the company has acquired Fluent (CAE), CFX (CFD), ICEM CFD Engineering (CFD), and most recently, Ansoft (electromagnetics and EDA). Interestingly, MSC.Software used to have an OEM agreement with one of Altair’s new partners, nCode, a leader in fatigue and durability analysis.
The open platform strategy is Altair’s effort to do an end run around their larger competitors. By aggregating software instead of companies, investments on both sides are relatively low. The ISV partners need to develop software that works under HyperWorks licensing, while Altair just has to manage the additional titles across their business. No money changes hands until customers start using the software.
Compared to ANSYS’ recent $800 million acquisition of Ansoft, the arrangement is much lower risk and lower cost. Humphrey, who was at SGI when they bought Cray Research in 1996, knows buyouts don’t always have a happy ending. “We’ve done some successful acquisitions here at Altair,” he says. “But anytime you acquire a company, there’s about a 50-50 chance of failure.”
But the open platform model itself is untested. Humphrey says Altair will make sure they have all the bugs worked out and the infrastructure in place before they expand the program further. And after they get some traction, they’ll start recruiting more partners into the program. According to him, they have a list of 30-35 ISVs that have expressed interest.
For now, they’ve probably picked all the low-hanging fruit with the initial six partners. Even though the software they’ve added is complementary to HyperWorks, Humphrey says the company is not ruling out bringing competitive software titles on board. If they go down that path, they’ll have to balance the impact on Altair’s internal R&D with the extra revenue they accrue from partner sales. From the corporate viewpoint, complete vendor neutrality is a good thing, since it does right by the customer and keeps money flowing into the company — which, in turn, could provide funding to retool native products. At some point though, your internal developers start to look like just another ISV partner.
“It’s an interesting debate,” admits Humphrey. “If we don’t have a best-in-class product in a particular segment, and we’re getting beat by a competitor in some scenarios, we can still provide the front end interface for that product. That’s a better position than losing the customer altogether.”